Learning a Controller for Soft Robotic Arms and Testing its Generalization to New Observations, Dynamics, and Tasks

Alessi, Carlo; Häuser, H.; Lucantonio, Alessandro; Falotico, Egidio

doi:10.1109/robosoft55895.2023.10121988

Cited by 5 publications

(1 citation statement)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[ 80 ] uses RL to make a soft robot able to work under different payload conditions. Going one step further, the authors of [ 81 ] proved the robustness of RL controllers when changing the geometry, velocity or materials of the robot.…”

Section: Related Workmentioning

confidence: 99%

Model-Free Control of a Soft Pneumatic Segment

García-Samartín,

Molina-Gómez,

Barrientos

2024

Biomimetics

View full text Add to dashboard Cite

Soft robotics faces challenges in attaining control methods that ensure precision from hard-to-model actuators and sensors. This study focuses on closed-chain control of a segment of PAUL, a modular pneumatic soft arm, using elastomeric-based resistive sensors with negative piezoresistive behaviour irrespective of ambient temperature. PAUL’s performance relies on bladder inflation and deflation times. The control approach employs two neural networks: the first translates position references into valve inflation times, and the second acts as a state observer to estimate bladder inflation times using sensor data. Following training, the system achieves position errors of 4.59 mm, surpassing the results of other soft robots presented in the literature. The study also explores system modularity by assessing performance under external loads from non-actuated segments.

show abstract

Section: Related Workmentioning

confidence: 99%

Model-Free Control of a Soft Pneumatic Segment

García-Samartín,

Molina-Gómez,

Barrientos

2024

Biomimetics

View full text Add to dashboard Cite

show abstract

Pushing with Soft Robotic Arms via Deep Reinforcement Learning

Alessi,

Bianchi,

Stano

et al. 2024

Advanced Intelligent Systems

View full text Add to dashboard Cite

Soft robots can adaptively interact with unstructured environments. However, nonlinear soft material properties challenge modeling and control. Learning‐based controllers that leverage efficient mechanical models are promising for solving complex interaction tasks. This article develops a closed‐loop pose/force controller for a dexterous soft manipulator enabling dynamic pushing tasks using deep reinforcement learning. Force tests investigate the mechanical properties of a soft robot module, resulting in orthogonal forces of N. Then, the policy is trained in simulation leveraging a dynamic Cosserat rod model of the soft robot. Domain randomization mitigate the sim‐to‐real gap while careful reward engineering induced pose and force control even without explicit force inputs. Despite the approximate simulation, the sim‐to‐real transfer achieved an average reaching distance of mm (), an average orientation error of rad () and applied pushing forces up to N. Such performance is reasonable for the intended assistive tasks of the manipulator. The experiments uncovered that the soft robot interacting with the environment exhibited torsional and counter‐balancing movements. Although not explicitly enforced, they emerged from the mechanical intelligence of the manipulator. The results demonstrate the potential of soft robotic manipulation via reinforcement learning.

show abstract

Learning Controllers for Continuum Soft Manipulators: Impact of Modeling and Looming Challenges

Falotico,

Donato,

Alessi

et al. 2024

Advanced Intelligent Systems

View full text Add to dashboard Cite

Soft manipulators, renowned for their compliance and adaptability, hold great promise in their ability to engage safely and effectively with intricate environments and delicate objects. Nonetheless, controlling these soft systems presents distinctive hurdles owing to their nonlinear behavior and complicated dynamics. Learning‐based controllers for continuum soft manipulators offer a viable alternative to model‐based approaches that may struggle to account for uncertainties and variability in soft materials, limiting their effectiveness in real‐world scenarios. Learning‐based controllers can be trained through experience, exploiting various forward models that differ in physical assumptions, accuracy, and computational cost. In this article, the key features of popular forward models, including geometrical, pseudo‐rigid, continuum mechanical, or learned, are first summarized. Then, a unique characterization of learning‐based policies, emphasizing the impact of forward models on the control problem and how the state of the art evolves, is offered. This leads to the presented perspectives outlining current challenges and future research trends for machine‐learning applications within soft robotics.

show abstract

Learning a Controller for Soft Robotic Arms and Testing its Generalization to New Observations, Dynamics, and Tasks

Cited by 5 publications

References 23 publications

Model-Free Control of a Soft Pneumatic Segment

Model-Free Control of a Soft Pneumatic Segment

Pushing with Soft Robotic Arms via Deep Reinforcement Learning

Learning Controllers for Continuum Soft Manipulators: Impact of Modeling and Looming Challenges

Contact Info

Product

Resources

About